446
PRANKHOVORKA AND WILLJAM C. DEARING
length 2537 b. In the presence of oxygen, this irradiation has been found to produce reaction such that on subsequent hydrolysis equal numbers of equivalents of chloride, acid and oxidizing agent are found: the reaction is assumed to be
[CONTRIBUTION FROM
Vol. 5;
+
2CCk 02 -+ 2COC12 4- 2C12. Using a monochloroacetic acid actinometer, quantum yields of one carbon tetrachloride oxidized per quantum absorbed have been found. PASADENA, CALIF.
RECEIVED JANUARY 2, I933
1
THE MORLEY C ~ I E M I LABORATORY, CAL WESTERN RESERVE UNIVER~ITY
The ‘‘Salt Error” and Normal Electrode Potential of the Quinhydrone Electrode at 25’ B Y FRANK HOVORKA AND WILLIAM
c. DEARING
The “salt error” of the quinhydrone electrode electrode for solutions of various salts, but with was recognized but neglected by Biilmann’ in the less precision due to uncertainties of the liquid initial study of the cell Pt/quinhydrone, H*(m), junctions involved. For further papers dealing Hz(l atm.)/Pt at 18 and 25’. From solubility with the quinhydrone electrode and related submeasurements Sorensen, Sorensen and Linder- jects the reader is referred to a bibliography comstrom-Lang2showed the “salt error” to be due to a piled by Morgan, Lammert and CampbelL’O In view of the increasing importance of the change in the activity ratio of hydroquinone and quinone, the dissociation products of quinhydrone. quinhydrone electrode in many fields it seemed From similar measurements Linder~trOm-Lang~ advisable to study some of the fundamental calculated the potential corrections for the quin- properties of the electrode in a more systematic hydrone electrode. He also determined by extra- manner and with greater precision than heretofore polation the value of E&, the potential of the attained. Foremost among the objectives was a quinhydrone-hydrogen cell at 18’ were water the study of the quinhydrone-hydrogen cell and a experimental liquid. Kruy t and Robinson4 con- determination of the “salt error” of the quinfirmed the results of Linderstrom-Lang regarding hydrone electrode in solutions of a large number the influence of salts on the solubilities of hydro- of electrolytes, each over such a range of concenquinone and quinone. Numerous factors which tration as to permit definite conclusions regarding influence the potential of the quinhydrone- the trend of the “salt error” and the possibility of hydrogen cell were carefully studied by Biilmann making adequate corrections. As a corollary of and J e n s a 6 Tammann and Jenckels and Cupr’ this was the evaluation of Eo,the normal electrode determined the “salt error” of the quinhydrone potential of the quinhydrone electrode. Several electrode in a limited number of solutions, but secondary objectives developed in the course of measurements were made against the calomel the investigation. These will be discussed later. electrode rather than against the hydrogen elecApparatus and Materials trode in the same solutions. Harned and Wrights Electrolytic hydrogen from a tank was washed with determined the normal electrode potential of the concentrated potassium hydroxide solution and water, quinhydrone electrode at intervals from 0 to 40’ then passed over platinum at 450”. Upon reaching the by an indirect method which required the com- thermostat the hydrogen passed through a washing bottle bining of data for two types of cells. In a similar containing the solution being studied and thence to the manner Urmbnczys determined the “salt error” electrode vessel which w a s similar in some respects to that used by Linderstrom-Lang.* It consisted of two tubes, of the quinhydrone electrode against the hydrogen C and D, of 40-cc. capacity, and E with a capacity of 20 (1) Biilmann, Ann. Chim., 11, 109 (1921). (2) Siirensen, Sdrensen and LinderstrBm-Lang, ibid., 16, 283 (1921). (3) Linderstrbm-Lang, Compt. rcnd. frau. lab. Carlsbcrg, IS, N o . 4 (1924); 16,No.3 (1925). (4) Kruyt and Robinson, Vcrslag. Akad. WelcnschaPPen Amslcrdarn, 81,812 (1926). (5) Biilmann and Jensen, Bull. soc. ckim.,41,151 (1927). (6) T a m m a n n and Jenckel, 2. anorg. allgem. Chsm., 171, 337 (1928). (7) Eupr. Pub. Facull8 Sci. Uniu. Masaryk, No. 133 (1931). (8) Harned and Wright, THISJOURNAL, SI, 4849 (19%). ‘(9) UrmLnczy, .tfaaivr C h r m i o i Folyoirnf, 39, 125 (1933).
cc. These were connected through the three-way stopcock, F. A heavy rubber cap, G, was fitted tightly over the bottom of the stopcock. This was sealed with lacquer and furnished adequate insulation from the bath. Two hydrogen electrodes of B. & S. No. 18 platinum wire and leads were sealed into a 5-mm. tube, leaving 17 mm. of the electrode wires exposed. This was fitted into a larger tube by means of a small rubber stopper. Two (10)Morgan, L a m B e r t and Campbell, Trans. Am. Ekchockcm.
.%.. 61, 406 (1932).
TKESALTERROR AND POTENTIAL OF THE QUINHYDRONE ELECTRODE
March, 1935
holes for the escape of hydropa were placed in this tube I cm. from the bottom. The exit tube, H, extended down the side of the thermostat. Two gold wire electrodes, L. & N. No. 7701, were mounted in a rubber stopper and occupied the tube D of the electrode vessel. One arm of a small stopcock also passed through the stopper to permit the equalization of pressure within and without. The electrode potentials were measured by means of a L. & N. type K potentiometer, type R galvanometer, and lamp and scale. The standard cell was of the unsaturated cadmium type, made by the Eppley Laboratories and certified by the Bureau of Standards. This was checked a t frequent intervals against another certified Eppley cell. The system was shielded and free from appreciable stray currents. The thermostat was held tit 25 * 0.01”. \ I
Preliminary work with the hydrogen electrode led to the following conclusions. (a) The deposition of gold plate prior t o platinizing does not improve the electrode. The removal (by electrolysis in hydrochloric acid) of the platinum black deposit on such a base has no advantage over the method finally adopted. (b) Deposits from pure platinic chloride are less satisfactory than those from solutions containing lead acetate. (c) Electrolysis in sodium hydroxide solution following the platinizing has no advantage over sulfuric acid alone. (d) Reasonably thin deposits of black are more satisfactory than thick when used in solutions where there is little danger of poisoning, and when the electrodes are replatidzed frequently. The procedure finally adopted was as follows. The old black was removed with a damp cloth and h e emery powder. This was followed by a similar polishing process employing alkaline scouring powder. Occasionally a final treatment in hot chromic acid or alcoholic potassium hydroxide was necessary to obtain even evolution of hydrogen when testing for cleanliness. The platinum black was deposited from a 2% solution of platinic chloride containing 0.02 g. of lead acetate per 100 cc. The electrodes were then electrolyzed in 7% sulfuric acid for about twenty minutes, rinsed thoroughly under the tap and with distilled water, and kept immersed in the latter when not in use. Procedures such as flaming, washing in absolute alcohol,
447
etc., were found t o be entirely unnecessary with the gold wire quinhydrone electrodes used. They were cleaned occasionally in hot chromic acid, and their behavior was a t all times quite regular. When not in use they were kept immersed in distilled water. There has been considerable diversity of opinion as to the best method for the preparation of quinhydrone. The product of the Eastman Kodak Company has been widely used and recommended by recent workers.loJ1 A supply obtained from that company was compared with another sample from the same company which had been in the laboratory for some time, and they were found to agree to within 0.00002 v. However, a later comparison with two samples prepared by the method of Biilmann and Lundl* and three samples prepared by the direct union of quinone and hydroquinone showed the Eastman product to give a distinctly negative potential with respect t o the others. The above five samples agreed among themselves t o within 0.00004 v., those prepared by direct union being more positive. Furthermore, the Eastman product after recrystallization from water gave a potential in excellent agreement with that prepared from quinone and hydroquinone. Equimolecular quantities of the compounds were dissolved in the minimum volumes of boiled distilled water a t 65’, filtered and the solutions mixed while warm. After cooling with ice the quinhydrone was atered on a Buchner funnel, washed four times with cold water and dried between filter papers a t room temperature. Constant boiling hydrochloric acid prepared according to Foulk and Hollingsworth’a constituted the stock acid solution and from this a 0.5 N secondary stock solution was prepared. The latter was used in making up all solutions which were 0.01 N with respect t o hydrochloric acid, the constant boiling acid being used for those of higher concentrations. Each solution was prepared by weighing out appropriate quantities of acid and salt and diluting in a calibrated flask at 25’. C. P. salts of the highest quality obtainable were employed, part being further purified by recrystallization. (It should be borne in mind that the measurements of the “salt error” which constitute this research are themselves differential in nature, and that any errors due to slight impurities are of second order. To illustrate: assuming that a sample of potassium chloride contains even 0.1% sodium chloride, it can be shown that for a 2 N potassium chloride solution, at which concentration the error would be greatest, this error is only 0.000001 volt, less than one-tenth the minimum experimental error.) The solutions were prepared from salts either in the anhydrous state or from concentrated stock solutions which were analyzed by standard quantitative procedures. The sulfuric acid solutions were likewise prepared by weighing appropriate quantities of an analyzed solution and diluting as stated above. Experimental Procedure The electrode vessel and electrodes were washed thoroughly with three portions of the solution being studied, (11) Morgan, Lammert and Campbell, Tars JOURNAL, 63, 454 (1931). (12) Biilmann and Lund, Ann. chim., 16,321 (1921). (13) Foolk and Hollingqworth, THISJOURNAL, 46, 1220 (1923)
448
~ A N HOVORW K AND
WILLIAM6.DEAJ~~NC
then solution was added until it stood a t J in all three arms. C and D each contained about 20 cc. The unit was lowered in the thermostat until the cross arms were just submerged. Hydrogen was introduced, and the rate so adjusted as to give about eight pulsations in the electrode tube every five seconds. These alternately bathed and exposed to hydrogen the upper centimeter of the electrodes. As soon as hydrogen was introduced the stopcocks F and I were opened for a moment to allow for slight readjustment of the solution level, then F was closed. The potential between the two hydrogen electrodes was then followed. This fluctuated widely for a short period but soon began to drop steadily. Within fifteen minutes the difference was usually
